60hp / flat torque curve remap for AMF

Robin_Cox

Member
A little knowledge is a dangerous thing. When paired with Ebay purchases and free dodgy software downloads the combination borders on insanity - but I hope that some folk may find the following informative. 18 months ago when I was having problems with my stock BHC car wheel-spinning on damp roads (crap budget front tyres when bought as well as urgently needed new front shocks) I was discussing remaps with Paul (@depronman of this parish) - who is knowledgeable in this area and asked a question about the possibility to knock up a map that would reduce likelihood of a Tdi breaking traction in slippery conditions - effectively a 'winter' mode. His considered response was that a torque limiter (fuel injection quantity vs rpm) tweak may well be the direction to go if that was the aim. Once I fitted decent tyres that proved to grip in 90% of the conditions encountered here in Scotland, I never went further with the discussion or idea last year but the seed was sown.

Having downloaded the various bits of software (MPPS v13 for transferring maps to and from the EDC15 Tdi ECU, and another one to open them, interpret and make modifications) - and bought the MPPS v13 ODB/USB dongle that allows a PC laptop to transfer maps to and from the ECU I started mucking about the other day with the AMF map from my project car. Again - thanks to Paul for a lot of advice on the following.

The EDC15 (Tdi ECU) has enough memory for two code blocks - ie the extensive sets of maps covering all parameters, normally allocated notionally for manual and automatic vehicles. As supplied, the AMF and BHC have both halves programmed identically, but the ECU is supplied programmed into "code block 2" for manual. And if untinkered with, so it stays.

However, with VCDS or some plug-in readers, it is possible to switch the car over to "code block 5" as if it was an automatic, and make use of the other programme. If you've tinkered with the other programme, then you have the option of swapping between the two relatively quickly (simpler than a full remap that takes probably 10 minutes in ideal circumstances getting equipment into the car and the laptop to speak to the car safely).

I took the original map for my car (downloaded from the vehicle), and having opened this up on the laptop, approached the torque limiter map in code block 5, and took the normal bell-curve-shaped injection quantity giving the normal surge of acceleration from 2000-3000rpm down to a flat line quantity of 35mg fuel per injection unit across the range.

This altered the predicted torque and power curves from the bell shape peaking over 200nM at 2300-2800rpm with power peaking at 82 (75 officially) hp we normally expect down to a ~140nM flatline from idle to red-line, and a linear power delivery curve peaking at 62hp at 3900.

I then uploaded this to the car, and initially ran it in the code block 2. Car ran completely normally (as one would hope). Then switched the car over to code block 5. Car started up completely normally. In local commuting the car drives completely normally part throttle at <40mph - as the driver wish (throttle) map mostly requests fuel injection quantities lower than the torque limiter graph until you get above 80% throttle input across the rev range whereupon the torque limit curve is the red line for injection quantity. The only difference you notice is that when accelerating up a slip-road in wet conditions now giving it some welly, it is a fair bit slower and a lot more linear in terms of gaining revs, and there is no torque-steer at all. Still able to run up to 60/70mph on the local dual carriageway - but obviously, accelerating into a space or another lane to effect a quick overtake is no longer a given. Reminds me of driving an old late 80s Transit 2.3 diesel minibus at University - it will get you there but it takes a bit more planning and looking at your mirrors.

This isn't the most exciting achievement in the remap world - I will openly admit. I'm waiting to see if this has any positive effect on fuel economy (unlikely as I only uploaded it half way through a month running an uprated AMF map to begin with and am also working on fuel gauge calibration - and I'm not Audi's technical centre who optimised the original maps as far as they could have done with the benefit of actually knowing what they are doing). However, I have now answered my initial question - whether it is possible to generate a map that mitigates the potential of the AMF / BHC to break traction in slippery conditions whilst maintaining basic driveability?

The answer is that it is, but I'm having to balance the satisfaction of achieving what I aimed to do with the tedium of turning an A2 into an 80s Transit as far as the fun quotient is concerned.
 

damadgeruk

Admin Team
Interesting, if only we could swap between blocks with the flick of a switch or by reading the key transponder. My first foray into ecu programming resulted in my having to remove the ECU to bench boot it to install a functional map. What are you using to mod?
 

AndyP

Member
What would probably be more effective would be to modify the traction control parameters, thereby not losing power in the dry.
 

Robin_Cox

Member
There was a discussion about how to write code to get an EDC15 to switch code banks using combinations of pedal / cruise control switch signals on the Nefmoto forum a couple of years ago that is way beyond my level of understanding unfortunately.

I suspect that playing with traction control settings (presumably part of the stability system?) would be rather more risky than turning down the wick - my car still does 80% of what my A2 does normally just without the big torque peak from 2000-3000rpm. This was purely an exercise in 'what if' ; another option could be to soften the fueling curve for the accelerator "driver wish" map so that you only get the full beans if you floor it intentionally, but that could lead to it being like a light switch and probably lead to a firey death...

As far as the mapping was concerned, it is a standalone non-Internet-connected windows 7 pc laptop that was rescued from a skip in work, given a deep clean and set up with only car software on it. MPPS v13 for the upload/download software and cable ; VAG-EDC15p to examine maps and tinker with them; also have VCDS lite and VAG Commander - which I have opened once with immense trepidation.

Day-in day-out I have an Xtools VAG 401 ODB reader that is used for reading/erasing faults, but also does some adaptations / coding including swapping ECU code blocks from 2 to 5 and back again more quickly than I can do with the laptop, have also been using it to adjust the fuel gauge as it was putting the "I'm going to run out of fuel" beep on with 10 litres in the 34 litre tank around 300 miles.
 

damadgeruk

Admin Team
Must look into the low fuel warning, I'm not a fan of the dash(and ear) invasion. I've been playing with EDC of late trying to disappear an EML. Keen to know if you continue testing, could be useful if we get more snow than of late(though I run all our cars on winter tyres, first set just fitted).
 

Robin_Cox

Member
I've got two complete sets of SE wheels with proper winter tyres in the garage that will go onto the everyday cars towards the end of the month having seen my other half unable to get off the end of the driveway in a 60hp Citigo during the Beast from the East aftermath 18 months ago.

Interestingly just refilled the 60hp remap car having completed 300 miles with the car now on a quarter tank indicated (3 weeks ago the fuel alarm went off at 305 miles with 27 litres needed to refill) ; today it took 25 litres. This is extremely unscientific : the first 200 miles was running a locally-sourced potent AMF remap that I won't talk about, and I was taking full advantage of it once I the car started breathing more freely. Then 100 miles substantially detuned as described above and I've driven pretty normally in this distance - basically not much difference in overall speed barring the occasional bursts of acceleration that the high map give you access to. There are arguments both ways on remaps and fuel efficiency. Owing to the increased flexibility of the 'improved compared to normal' map - one can use a higher gear trundling around which presumably uses less fuel due to the lower revs - and then the flip side, the restricted map I have been trialling is a bit of a chastity belt when it comes to using the loud pedal.

If I don't die of boredom using my own map I might have some more reliable statistics on fuel in around 3 and a half weeks.
 

AndyP

Member
If you really want to save fuel trip the ECU into limp-home mode so there is no boost - the mpg is impressive! Really tedious going up hills though ... :)
 

Un4tural

Member
If you really want to save fuel trip the ECU into limp-home mode so there is no boost - the mpg is impressive! Really tedious going up hills though ... :)
Hah was annoying 100miles when i had the problem! Returned 80+ MPG though - not sure why it cant dial down the turbo to cruise and get that mpg without limp mode. Struggled to get to 70mph badly. might be a thought for a map though to adjust - was always curious to fiddle with the maps and such.
 

Robin_Cox

Member
I've not yet played with the boost map - it would be easy to limit it to a 'low pressure' mode but to be honest it seems to be working fine at the moment with what it has got.
 

Robin_Cox

Member
Today's refill after a couple of weeks of pure commuter driving on the low power map resulted in a step up from 55mpg to >58mpg (~308 miles on 24 litres) with no hint of wheelspin even on properly sodden roads - bear in mind that the car hasn't reached 90 degrees on most occasions as it is only 7-8 miles so it doesn't reflect the mpg I would expect on a longer drive or in warmer months (temperature has dropped off here since September). One morning the Webasto kicked in as well. Longer drives do get it up to a stable 90 degrees (I put a new thermostat in during the rebuild). A couple of times I wouldn't have minded a wee bit more poke (uphill sliproads or when wanting to accelerate into a fast lane gap to overtake can take a bit of planning), but the map has not been the hair shirt one would assume it might be and isn't really all that slow (unless you compare it to the torque curve of the normal AMF or various AMF+ maps that give it a bit to a lot more puff). It will still achieve 70mph on the dual carriageway, although is happier around 55-60. The main thing wasn't trying to turn an AMF into an ANY - it was to have the option to remove the unexpected from the performance repertoire in icy conditions - and on that basis I am satisfied.
 

Kernow A2

Active Member
Has anyone thought of looking at the tuning units the Jap tuning crowd use to control boost.... Apexi springs to mind.....whereas they want to gain boost I'm sure they could be used to limit boost without having to resort to a remap
 

Vantagemech

Member
Surely the way to go would be bring the TC in earlier? There must be some parameters that could be adjusted to do this? On the race cars they have a Bosch system that has both adjustable TC & ABS settings. Im not saying you can do it but by perhaps reducing the slip figure between wheels could make it more intrusive.
I know from running one of Pauls maps (17k with zero issues) that even in dty conditions you can light up the TC when a little over enthusiastic on the loud pedal...
 

Robin_Cox

Member
As stated above, this was an experiment using the tools I had available - reducing the overall fuelling map is simple and direct, and all it has done is decapitate the torque curve peak. Adjusting traction control settings is a different beast altogether given the interaction with stability control and ABS etc.; way outwith my capabilities.
In my experience of the traction control kicking in in deep snow 15 months ago (on new winter tyres), having it cut off the power completely when I was generating momentum to rock the car out of the rut it was in was more hassle than it was worth again because it was a very abrupt change that was no longer proportional to what my pedal inputs were requesting - ie the same but opposite problem to generating unexpected and sudden wheelspin when trying to be calm and considered with pedal inputs. Bringing in this effect earlier would to my mind be counterproductive.

An easy answer would be 'learn to drive'. Having experienced how easy a 2CV on normal tyres is to drive in deep snow more than 20 years ago as a student, a rally driving friend of my parents pointed out how it was ideal *because* of the fact it had next to no torque to catch you out, narrow tyres, large suspension travel and weighed next to nothing, so if I've been able to reproduce the first characteristic that is something.
 
Top